Presentation is loading. Please wait.

Presentation is loading. Please wait.

The Battle for Accountable Voting Systems Prof. David L. Dill Department of Computer Science Stanford University

Similar presentations

Presentation on theme: "The Battle for Accountable Voting Systems Prof. David L. Dill Department of Computer Science Stanford University"— Presentation transcript:

1 The Battle for Accountable Voting Systems Prof. David L. Dill Department of Computer Science Stanford University

2 Outline Principles Trust and DREs Voter verifiable audit trail Conclusion

3 Role of Elections Democracy depends on everyone, especially the losers, accepting the results of elections. “The people have spoken... the bastards!” - Dick Tuck concession speech

4 Burden of Proof We should be able to prove that elections are accurate. –Procedures and equipment must be reliable and secure. –Election results are routinely and meaningfully audited. Audit: independently reconstruct election results from the original records. With conventional paper-based systems, manual recounts.

5 Integrity With Paper Ballots Integrity measures (with good procedures). –Voter makes a permanent record of vote. –Locked ballot box is in public view. –Transportation and counting of ballots are observed by political parties and election officials. Everyone understands physical security of paper ballots. Any new system should be at least this trustworthy.

6 Trust “You have to trust somebody.” We only need to trust groups of people with diverse interests (e.g., observers from different political parties).

7 Outline Principles DREs Voter verifiable audit trail Conclusion

8 DRE Definition DRE = “Direct Recording Electronic” For this talk, “DRE” does not include machines with voter verifiable paper records.

9 The Man Behind the Curtain Suppose voting booth has a man behind a curtain –Voter is anonymous –Voter dictates votes to scribe. –Voter never sees ballot. There is no accountability in this system! (analogy due to Dan Wallach and Drew Dean)

10 The DRE Auditing Gap Screen touches Recorded votes DRE System ? President Joe Blow President Fred Derf Any accidental or deliberate flaw in recording mechanism can compromise the election.... Undetectably!

11 Integrity of DRE Implementations Paperless electronic voting requires DRE software and hardware to be perfect. It must never lose or change votes. Current computer technology isn’t up to the task. ? ?

12 Program bugs We don’t know how to eliminate program bugs. Inspection and testing catch the easy problems. Only the really nasty ones remain –obscure –happen unpredictably.

13 Security Risk What assets are being protected? –At the national level, trillions of dollars. Who are potential attackers? –Hackers, Candidates, Zealots, –Foreign governments, Criminal organizations Attackers may be very sophisticated and/or well-financed.

14 A Generic Attack Programmer, system administrator, or janitor adds hidden vote-changing code. Code can be concealed from inspection in hundreds of ways. Code can be triggered only during real election –Using “cues” - date, voter behavior –Explicitly by voter, poll worker, or wireless network. Change small % of votes in plausible ways.

15 Generic attack DREs are creating new kinds of risks. Nationwide fraud becomes easier than local fraud. Local election officials can’t stop it!

16 Threats From Insiders FBI: “The disgruntled insider is a principal source of computer crimes.” – The 1999 Computer Security Institute/FBI report notes that 55% of respondents reported malicious activity by insiders. Crimes are easier for insiders (e.g., embezzling).

17 Voting is Especially Hard Unlike almost every other secure system, voting must discard vital information: the connection between the voter and the vote.

18 Comparison with banking Electronic audit records have names of everyone involved in every transaction. Banks usually have paper backup!... And computer crime still occurs -- especially by insiders. but Fraud can be quantified (we can tell when it happens). Customers are protected.

19 What software are we running? We cannot verify that desired software is running on a computer. Stringent software design/review (even formal verification) doesn’t solve the problem. Open source does not solve the problem. –“Disclosed” source is, however, highly desireable!

20 Summary of Technical Barriers It is currently impossible to create trustworthy DREs because: We cannot eliminate program bugs. We cannot guarantee program security. We cannot verify that the desired software is running on the computer.

21 Outline Principles Trust and DREs Voter verifiable audit trail Conclusion

22 The Man Behind the Curtain Now, suppose the man who filled out the ballot –Shows you the ballot so you can make sure it is correct. –Lets you put it in the ballot box (or lets you watch him do it). There is accountability –You can make him redo the ballot if it’s wrong. –He can be fired or arrested if he does it wrong.

23 Voter Verifiable Audit Trail Voter must be able to verify the permanent record of his or her vote (i.e., ballot). Ballot is deposited in a secure ballot box. –Voter can’t keep it because of possible vote selling. Voter verified records must be audited, and must take precedence over other counts. This closes the auditing gap.

24 VVAT is not enough Closing the audit gap is necessary but not sufficient. Additional conditions: –Physical security of ballots through final count must be maintained. –Process must be transparent (observers with diverse interests must be permitted at all points). There are many other requirements, e.g., accessibility.

25 Manual Recounts Computer counts cannot be trusted. Like other audits, independent recounts should be performed at least –When there are doubts about the election –When candidates challenge –On a random basis Computer-generated ballots can have additional security features. –Digital signatures/time stamps –Matching identifiers for reconciling with paper ballots.

26 Options for Voter Verifiable Audit Trails Manual ballots with manual counts. Optically scanned paper ballots. –Precinct-based optical scan ballots have low voter error rates. Touch screen machines with voter verifiable printers. Other possibilities (unproven! ). –Other media than paper? –Cryptographic schemes? For now, paper is the only option.

27 Outline Principles Trust and DREs Voter verifiable audit trail Conclusion

28 Key points Election equipment should be proved reliable and secure before it is deployed. There is little evidence that DREs are safe, and a lot of evidence to the contrary. The problems cannot be fixed without a voter verifiable audit trail of some kind. With a voter verifiable audit trail and due attention to election practices, the problem can be solved.

29 The Big Risk All elections conducted on DREs are open to question.

30 More information is available at our website., Inc. is now a 501(c)(4) non-profit.

31 Major Events Since Jan 2003 Jan, 2003. “Resolution on Electronic Voting” finalized and signed by 3 people. Jan 2003. Santa Clara County (CA) Recommends Buying DREs. Computer Scientists Speak Out. Feb 2003. CA Ad Hoc Task Force on Touch- Screen Voting Convened. ? Feb/Mar 2003. Rush Holt Introduceds HR 2239 -- “Voter Confidence and Increased Accessibility Act” Requiring a Voter Verifiable Paper Trail. May 2003. Task Force Recommends “Voter Verifiable Audit Trail” by 2010.

32 Major Events Since Jan 2003 June, 2003. CA Secretary of State Kevin Shelley receives 6,000 letters -- 4,000 in favor of a voter verifiable paper trail. July, 2003: Johns Hopkins/Rice Report finds serious security problems with Diebold software Nov 2003: CA SoS Shelley announces paper trail requirement for California (2005/2006) Jan 2004: SERVE program cancelled. Mar 2004: Various machine failures in primaries

33 Mock Election Friday before the panel on e-voting.

Download ppt "The Battle for Accountable Voting Systems Prof. David L. Dill Department of Computer Science Stanford University"

Similar presentations

Ads by Google